Brian W. Pence, PhD MPH Associate Professor of Community and Family Medicine and Global Health Duke University June 4, 2012

Changes in HIV Providers’ Management of Depression after

Integration of Treatment Support into Clinical Care

Thank you to…

Abstract lead author Kiana Bess

SLAM DUNC Study Team:

Duke: Julie Adams; Jordan Akerley; Kiana Bess; Jacob Kirkorowics; Kathryn Schley; Nathan Thielman; Quinn Williams; Abigail Zeveloff

UAB: Michael Mugavero; Teena McGuinness; Riddhi Modi; James Raper; James Willig: Anne Zinski

UNC: E. Byrd Quinlivan; Malaika Edwards; Bradley Gaynes; Amy Heine; Tiphany Jackson; Julie O’Donnell; Katya Roytburd

Funding: NIMH; NINR

Duke, UNC & UAB HIV Clinics and patients

The “HIV Treatment Cascade”

Gardner et al. CID 2011;52:793-800.

Depression and the HIV treatment continuum

HIV

Infection

HIV

diagnosis

Engagement

and retention

in care

ART

initiation

Adherence

and virologic

suppression

Depression inversely associated with…

Pence et al. AIDS 2012;26(5):656-8

Some evidence that depression treatment can improve ARV

adherence and HIV clinical outcomes.

The depression treatment continuum

Pence et al. AIDS 2012;26(5):656-8

Prevalent

depression

Clinical

recognition

Initiation of

treatment

Adequacy

of

treatment

Remission

of

depression

~50% of cases of

depression go

unrecognized in HIV

clinical care

~50% of those

recognized

never start

treatment

• Counseling requires a minimum contact

intensity.

• Pharmacotherapy requires aggressive dose

adjustment based on response.

• Perhaps one-third of those being treated are

receiving adequate treatment.

If adequate treatment

is provided, response

rates are high

(~70%).

The “Depression Treatment Cascade” for HIV Patients

0%

20%

40%

60%

80%

100%

Of p

reva

lent c

ases…

Rec

ognized

clinica

lly

Any tre

atment

Adequate

trea

tment

Achie

ved re

miss

ion

Screening

Depression Care

Manager support

Goal:

Pence et al. AIDS 2012;26(5):656-8

| 95% Bayesian Credible Interval

SLAM DUNC Study

Strategies to Link Antidepressant and Antiretroviral Management at Duke, UAB, and UNC

NIMH-funded R01, 2009-2014

Randomized controlled trial to test the effect of depression treatment on ARV adherence

Sites: HIV clinics at 3 academic medical centers

Pence et al. Contemporary Clinical Trials 2012; 33(4):828-38

SLAM DUNC Study

Population: HIV clinic attendees with current major depression

Intervention: Measurement-Based Care depression treatment

Depression Care Managers provide decision support to HIV providers for antidepressant prescription and management

Comparison: Usual care

Outcome: ARV adherence Pence et al. Contemporary Clinical Trials 2012; 33(4):828-38

Aim of this presentation

Does Depression Care Manager’s provision

of decision support for intervention-arm study

participants change HIV providers’

depression management practices for all

patients?

Methods

Semi-structured in-person interviews with HIV providers

Before study launch

One year after study launch

Topics

Perception of prevalence of depression and depression treatment in caseload

Confidence treating depression

Specific depression treatment practices

Methods

9 “best-practices” principles of depression treatment

defined from national guidelines

American Psychiatric Association* (for all patients)

NY State Dept of Health** (for HIV patients)

Providers’ self-reported practices compared to best-

practices principles and assigned a numeric score

reflecting quality relative to each principle

Summary best-practices score created using factor

analysis from item-specific scores * American Psychiatric Association. Practice Guideline for the Treatment of Patients With Major Depressive

Disorder, Third Ed. Washington, DC: American Psychiatric Association Press; 2010.

** New York State Department of Health. Depression and mania in patients with HIV/AIDS. New York: New

York State Department of Health, 2010.

Best-Practices Principles

1. Routine screening with standardized measure

2. Assess need for treatment based on severity

3. Base treatment choice on history and severity

4. Adjust antidepressant starting dose for ARV interaction

5. Follow-up new antidepressant prescription in 4 weeks

6. Assess effectiveness with standardized symptom severity

measure

7. Increase dose based on symptom severity and drug

tolerability

8. Utilize full FDA dosing range

9. Switch antidepressants based on formal assessment of

response, tolerability, and adequacy of trial

Screening

Initiation

Management

Participants (n=41* with baseline and 1-year follow-up interviews)

Baseline Characteristic N (%) or Median (IQR)

Clinical training

MD-Attending

MD-Fellow

Nurse Practitioner

Physician Assistant

21 (51%)

13 (32%)

3 (7%)

4 (10%)

Clinical effort

% clinical effort on HIV care

30% (20-50%)

80% (60-95%)

Years providing HIV care

<5 years providing HIV care

9 (3-20)

15 (37%)

Very or extremely confident…

Prescribing an initial antidepressant

Switching or augmenting antidepressants

Depression treatment is part of role

25 (63%)

5 (13%)

29 (71%)

* All providers were interviewed at baseline (n=48). At one year, 3 MD-fellows had left; 41 of 45 remaining providers

completed the one-year interview.

Baseline indicators of depression treatment quality

Bess et al. IAPAC 2011 abstract # 70029.

Change in self-reported depression treatment practices after 1 year

Baseline 1 year

+37% *

-4%

-14%

+16%

+2%

+13%

-29% *

+22%

-6%

* p < 0.05

Depression treatment best-practices summary score

5.7

(95% CI :4.9-6.4)

5.9

(95% CI :5.3-6.6)

p > 0.05

Mo

re g

uid

elin

e-

co

nco

rdan

t p

ractice

“How has your depression treatment practice changed in the past year?”

Open-ended question

Most common response: Provision of routine screening has

raised awareness of depression

and likelihood of discussing depression during clinical

encounter

Next most common response: Practice has not changed

Rare responses: Concrete practice changes

Follow up sooner after starting AD (n=1)

More likely to increase AD dose (n=3)

More likely to change AD (n=1)

“How has your depression treatment practice changed in the past year?”

“[I am] more comfortable addressing [depression] and

providing treatment with the care manager.”

“I didn’t feel it was my role at all … as a physician assistant

[and] in Infectious Diseases … I didn’t feel that the

physicians really thought it was much within their scope

either … but now there is more support, I have learned more

about the drugs, [and] I’m feeling more able.”

“I haven’t changed my practice, other than that I listen to

you guys.”

Interpretation

Providers have responded very positively to the

provision of routine depression screening in clinic

After 1 year of Depression Care Manager decision

support for intervention patients, providers report

relatively little change in practices related to the

initiation or management of antidepressant treatment

in all patients (without DCM support)

In contrast, adherence to guidelines in intervention

patients with direct DCM support is very high

Role of Depression Care Manager

Interpretation

Presence of study has raised awareness but has not

changed ongoing depression management practices

Supports need for measurement-based depression

treatment decision support for all patients

No evidence of contamination (decision support

around best-practices depression treatment appears

not to have generalized beyond intervention patients

Limitations

Self-reported practices

Providers may have “learned the right answer” after

1 year without having changed practice

Caution should be used in interpreting numeric

scores derived from qualitative data

Sites are academic medical centers with research-

oriented clinicians – practices may differ elsewhere

Strengths

Response rate: >90% of HIV providers at 2 large

academic medical centers

Detailed qualitative assessment of provider

practices

Very little published data on depression treatment

practices of HIV providers

Is this a fair expectation of HIV providers?

High burden and negative consequences of

depression

Rise of “medical home” model of care

Availability of evidence-based decision support

models (Measurement-Based Care)

Depression treatment summary score, by clinical training

Depression treatment summary score, by clinical training

Depression treatment summary score, by years of clinical experience

Depression treatment summary score, by years of clinical experience